Research Of P2P Media Streaming Peer Dispatching Based On Incentive Mechanism

P2P system develops fast due to its system characteristic, decentralizedarchitecture. Along with the popularization of network technology, P2P media systemsoars fast. Traditional online media system has been replaced. P2P media becomesmain stream. P2P media system resolves the problem, which tradition service needslarge scale servers, high bandwidth, and low performance in heavy concurrencesituation. P2P media system moves much of system load to its user by utilizing theiroften leisure, bandwidth. This strategy makes system user be consumer and producerof system resources. However, not all system users are so willing to contributethemselves. Selfish actions, like free-riding, are common in system. Such selfishactions contradict the system goal and compromise performance since load is heavyon other peer while selfish peer has no or little load, making advantage of P2Parchitecture be limited. Simultaneously, dispatching algorithm is not fair in old P2Psystem. It aims at high throughput, causing dispatching fairness and high switchinglatency problem.In response to these issues, this paper proposes a new P2P streaming systems,and describes and analysis its overall architecture and node modules. Specially, basedon the studying and comparing the existing incentives and node scheduling method,the author absorbs the advantages of the existing mechanisms and then proposes ascore incentives method. According to the upload bandwidth, online duration, uploadtraffic statistics, and other parameters of nodes, the nodes are divided into two groups.According to the integral strategy, the system provides the good nodes with thehighest quality services.The author proposes an improved P2P media streaming peer dispatching methodbased on score mechanism. This method, according to the downloads, contributionsand some other features of the nodes, packet loss rate for example, scores andschedules the nodes, and by the way of the prior nodes first reduces their handoverlatency and improve the users’ experience. In the end, analysis and compared theexisting P2P network simulation tools, the paper uses the most appropriate PeerSim asthe simulation tool. Through the experimental analysis comparing, the paperconcludes the incentive mechanism and node dispatching mechanism having some certain advantages and running well. In the real environment, needing to considermore factors, so, there are still a lot of works need to be done, such as the conspiracyissue in the process of scoring.